The present invention generally relates to a method for tank leak diagnosis.
One of such methods for a tank leak diagnosis is disclosed for example in U.S. patent document 2002/0139173A1. The tank leak diagnosis in accordance with this method is performed after turning off of an internal combustion engine. A pump, after turning off of the internal combustion engine, generates a negative pressure in the fuel tank. A leak in the fuel tank is recognized by a pressure increase in the fuel tank, which is caused by the leak. This method however has the disadvantage in that the negative pressure is produced after the turning off of the internal combustion engine, since the gas which is aspirated from the tank ventilating device can no longer be supplied for combustion in the internal combustion engine. Instead, the gas must be intermediately stored in an additional storage which is quite complicated and expensive and not discharged into the atmosphere. Moreover, it is also disadvantageous that the negative pressure is produced by an additional pump.
German patent document DE 198 20 234 C2 discloses a method of a tank leak diagnosis, in which during approximately the whole operation of the internal combustion engine a constant negative pressure in the fuel tank is maintained. This method has the disadvantage that a storage which receives the evaporated fuel can not be rinsed to a full degree, when the storage is held approximately during the whole operation of the internal combustion engine under a predetermined negative pressure. Thereby the storage, during turning off of the internal combustion engine, is not completely emptied, so that the storage capacity of the storage and a loading time during which fuel can be received in the storage in a loading phase is reduced when compared to a completely regenerated storage.